- Email: [email protected]
A deletion mutation at the 5′ end of Escherichia coli 16S ribosomal RNA
98
Biochimica et Biophysica Acta, 1050 (1990) 98-103 Elsevier
BBAEXP 92098
A deletion mutation at the 5' end of Escherichia coli 16S ribosomal RNA Pierre Melan~on, Daniel Leclerc and Lra Brakier-Gingras D~partement de Biochimie, Universit~de Montreal, Montreal, Quebec (Canada) (Received 16 May 1990)
Key words: ribosomal RNA; 16S; Translation initiation; Translation accuracy
A deletion of five nucleotides was introduced at the 5' end of the Escherichia coli 16S rRNA gene cloned in an appropriate vector under control of a T7 promoter. The 16S rRNA generated by in vitro transcription could be assembled into 30S subunits. The deletion did not affect the efficiency of translation of natural messengers and the correct selection of the reading frame. However, it reduced the binding of the messengers, which suggests that the 5' end of 16S rRNA is located on the pathway followed by the messengers on the 30S subunits. The deletion also restricted the stimulation of misreading by streptomycin in a poly(U)-directed system. This is in accord with the proximity of the 5' end of 168 rRNA to proteins 84, $5 and S12, which are known to be involved in the control of translational accuracy.
Introduction The process of translation initiation involves the selection by the 30S ribosomal subunits of an initiator codon on the messenger RNA. This codon is AUG in about 90% of Escherichia coli genes. A critical determinant for the efficiency of the initiation step is the lack of secondary structure in the mRNA around the initiator codon. Various interactions have been proposed to stabilize the binding of the messenger to the 30S subunits. The best-characterized is the interaction between the Shine-Dalgarno sequence, a purine-rich sequence on mRNA, upstream of the initiator AUG codon, and the anti-Shine-Dalgarno region, a complementary pyrimidine-rich sequence near the 3' end of 16S rRNA (reviewed in Refs. 1-4). The 5' end region of 16S rRNA (Fig. 1) is a likely candidate for an additional interaction with sequences downstream from the initiator codon, at the beginning of the coding region of mRNA [7,8]. The 5' end of 16S rRNA has been located at the rear of the 30S subunit, below the 530 loop (Fig. 2). It is distal from the decoding region in the cleft, and is proximal to proteins $4, $5 and S12. Affinity labeling studies (Refs. 20-22 and reviewed in Ref. 23) have shown that proteins $4, $5 and S12 can be
Correspondence: L. Brakier-Gingras, Drpartement de Biochimie, Universit6 de Montrral, Montrral H3C 3J7, Canada.
linked to the messengers. Moreover, crosslinking and footprinting studies demonstrate a direct interaction between the 5' end of 16S rRNA and proteins $4 and $5 [24,25]. These observations support an interaction between the 5' end of 16S rRNA and mRNA. Proteins $4, $5 and S12 are also known to be involved in the control of translational accuracy and in the response to streptomycin, an error-promoting aminoglycoside antibiotic (reviewed in Refs. 26-28). Their proximity to the 5' end region of 16S rRNA also suggests that this region could play a role in the control of translational fidelity, in addition to interacting with mRNA. We and others have recently developed a method which functions entirely in vitro to investigate the consequences of site-directed mutagenesis in E. coli 16S rRNA on ribosome function (Refs. 29-31 and reviewed in Ref. 32). In this system, mutations are introduced into the 16S rRNA gene cloned in an appropriate vector, directly downstream from a "1"7 promoter and flanked at its 3' end by an appropriate restriction site. In vitro transcribed mutated 16S rRNA is then assembled into 30S subunits, which are assayed for their protein synthesis activity. Using this system, we have previously shown that suppressing the Shine-Dalgarno interaction by deleting the anti-Shine-Dalgarno region at the 3' end of 16S rRNA neither affected the capacity of the 30S subunits to translate mRNA nor their aptitude to select the translation start sites correctly [33]. This was in full agreement with a study by Calogero et
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
99 530 LOOP
Materials and Methods
,°CC°cG 530
C (3
520-ACc
G/
G-C G-C CU GU
A AU GU__A. U
5'ko c - o . \ U--A
_ \
~ k 31~ .
u~Ac-c-54o cGG,,-~
~A G
~-~ A -- U 550
u',% A c-G
B901
uAC GGGA~uACGGccGc 2
o ~ ~ ""U / G 920
9 1 5 AREA G =~,v
c uCGG.AC'AA~C^CaO~Q~---~ .~^C.,,.,U..4" """'"
,~G A A
\uU~u u
~;AGu UGa \
^1o
C
ACUGUUCCG~ A
A
UU" AG--C G ~ A^',, A .... l A ^C ~,,A t, 30 , G4-.-~ ~UAt; 500 . t , r_t, U/
C C~1¢00 ,.,
'~'~'AUGUC C U G~ 400~C~ \_G G CU, GC~40
A A5 A I AJ
t, C C G
5'
u
/
:c
A
-®u
A A
C G
C--G U 14.10---A -- U-1490 A C --G A-15OO CmG C
®'@
U.G G.u G--C
A
A G
Fig. 1. Secondary structure of a portion of E. coli 165 rRNA [51, showing the 5' terminal region and parts of the different domains. The extent of the deletion introduced at the 5' end is indicated by the arrows. The circled nucleotides are protected against chemical probes by streptomycin [61.
al. [34], who did not find any difference when comparing the translation of two mRNA, which were identical except for the presence or absence of the Shine-Dalgarno sequence. In the present study, we have investigated how deleting five bases at the 5' end of 16S rRNA affects the activity of ribosomes.
Cleft
(decoding center)\c--7
/
Fig. 2. Model of the 305 sub~_it as seen from the solvent side [91. The positions of the 530 loop [10-121, the 5' end of the 16S rRNA [131, and the decoding center [14,15] have been mapped by immune and DNA-hybddizafion electron microscopy. Proteins S4, $5, $7 and S12 have been mapped by neutron diffraction [16] and immune electron microscopy [9,17]. The dotted area indicates the position of the 5' terminal and the pseudoknot helices of 16S rRNA at the interface with the 50S subunit [5,18,19].
Construction of the transcription vectors and preparation of the 16S rRNA transcripts The transcription vector pDI.A417 is a derivative of the phagemid Bluescript SK- (Stratagene). A first derivative of Bluescript (BLM1) was constructed by creating a NdeI site, directly upstream from the T7 promoter. The phagemid pDL4417 was then obtained by inserting between this Nde I site and the Xba I site in the polylinker, an NdeI-XbaI fragment from p P M l l 4 [31], which contains the E. coli 16S rRNA gene preceded by a T7 promoter. Two GC pairs have been inserted between the end of the T7 promoter and the beginning of the 16S rRNA gene. The 3' end of the 16S rRNA gene is part of a Bsu36I site, immediately followed by a XbaI site. A deletion mutant of pDL4417, pPM5'-A5, lacking 5 nucleotides at the 5' end of the 16S rRNA gene, was obtained by the phosphorothioate method of Eckstein [35,36], using the Amersham in vitro oligonucleotide-directed mutagenesis system. In vitro transcription of pDL4417 or its deletion derivative, linearized with Bsu36I (New England Biolabs), was carried out following standard procedures [37,38]. As a consequence of the construction of the plasmids, all 16S rRNA transcripts contain two additional GC residues at the 5' end. The 3' end is identical to that of natural 16S rRNA [31]. Assembly of 30S subunits The 30S subunits were assembled with either the full-length 16S rRNA transcript or its deletion derivative, as described previously [31,33]. The 5' end of 16S rRNA was analyzed by primer extension dideoxy sequencing [39] after phenol-extraction of the RNA from the reassembled subunits. Primer extension was done with reverse transcriptase (Pharmacia), using a (5'32p)-labeled synthetic deoxynucleotide primer [40], complementary to bases 31-45 in the natural 16S rRNA. Primer extension was also carded out without the dideoxy nucleotides in order to assess the amount of contaminating natural 16S rRNA in the reconstituted 30S subunits. Protein synthesis assays Polypeptide synthesis was programmed either with MS2 RNA or with the late mRNAs from the T7 phage. The reaction was carried out as described [31,33], using 0.16 A260 unit of 30S subunits and 0.33 A260 unit of 50S subunits. The translation products were analyzed by one-dimensional polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate as described [33]. Misreading assays under the direction of poly(U) were as described [31]: the cognate amino acid, [12C]phenylalanine, and the non-cognate amino acid, [3H]isoleucine (52 Ci/mmol), were added simulta-
100 neously at of concentrations of 10 mM and 1 mM, respectively. Streptomycin was added at a molar ratio of 5 per ribosome. In parallel assays, the incorporation of the cognate amino acid, [3H]phenylalanine (5 Ci/mmol), was measured in the presence of [12Clisoleucine and streptomycin. Binding of messenger RNA to 30S subunits The late mRNAs from the T7 phage were labeled with 35S during in vitro transcription, by including 25 #Ci of [a-3~S]UTP (Amersham; 1250 Ci/mmol) to 1 ml of the transcription mixture. Binding assays were performed by Millipore filtration, using 0.16 or 0.32 A260 unit of 30S subunits and 1 A260 unit of T7 messengers. The complex between the subunits and the mRNA, (200 #1), was incubated for 15 min at 37°C in a 50 mM Tris-HCl buffer (pH 7.8) containing 7.5 mM magnesium acetate, 100 mM NH4C1 and 1 mM dithiothreitol, then diluted to 3 ml with the same buffer, filtered through Millipore nitrocellulose filters (HA-0.22 m#) and washed three times with 3 ml of a cold buffer containing 10 mM Tris-HCl (pH 7.5), 10 mM magnesium acetate, 60 mM NH4C1 and 5 mM fl-mercaptoethanol. Filter-bound 355 w a s measured by liquid scintillation counting. Values were corrected for background retention of mRNA in the absence of 30S subunits. The specificity of the binding of mRNA was assessed by verifying the absence of retention with 50S subunits. Results
The 30S subunits were assembled with the full-length 16S rRNA transcript or its deletion derivative lacking five nucleotides at the 5' end. The reconstituted 30S particles comigrated with native 30S subunits in a sucrose gradient and analysis by two-dimensional gel electrophoresis indicated that they contained the full complement of ribosomal proteins. RNA sequencing analysis confirmed that the 5' end of the 16S rRNA transcripts had the expected sequence and primer extension assays indicated that the transcripts represented 95% or more of the 16S rRNA in the reconstituted subunits, the remaining 16S rRNA being natural 16S rRNA brought in as a contaminant of the ribosomal proteins (data not shown). The deletion at the 5' end of the 16S rRNA did not significantly affect the capacity of the 30S subunits to translate natural mRNAs, such as MS2 RNA or T7 phage late mRNAs (Table I). Furthermore, kinetic studies revealed no difference between mutated and unmutated 30S subunits during the time course of the reaction (not shown). The proteins synthesized by the mutated and m u t a t e d ribosomes were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis. The patterns of the trallslation products were identical whether the 30S subunits contained the full-length or the mutant 16S rRNAs (Fig. 3): a single
TABLE I Protein synthesis activity of reconstituted 30S subunits The activity of reconstituted 30S subunits was assessed by measuring the incorporation of [3H]leucine (1.9 Ci/mmol) and [3Hlvaline (1.8 Ci/mmol) into trichloroacetic acid-insoluble material under the direction of MS2 RNA or T7 late mRNAs. Each sample (100 ~1) contained 0.16 A260 unit of 30S subunits and 0.33 A260 unit of 50S subunits. Incubation was for 60 min at 37°C. Results are the means of four assays with independent preparations of 30S subunits. Standard deviation of the means was less than or equal to 17%. Blanks without 30S subunits have been subtracted. Origin of the 16S rRNA
Activity (in clam) with MS2 RlqA
with T7 late mRNAs
Natural 16S rRNA (isolated from native 30S subunits)
8 648
12 490
Synthetic full-length 16S rRNA
7160
10208
Synthetic truncated 16S rRNA (-5 nucleotides)
6 668
9 346
band corresponding to the coat protein was seen in all cases with MS2 RNA. A more complex pattern was observed with T7 phage late mRNAs, but the sizes and relative amounts of the synthesized proteins were identical with all types of 30S subunit. These results indicate that the deletion did not prevent the correct selection of the reading frame of translation and that the 5' end region of 16S rRNA does not appear essential for protein synthesis. However, the binding of mRNA to the mutated 30S subunits lacking five bases at the 5' end of 16S rRNA was decreased, as shown by the filtration assays measuring the binding of labeled T7 late mRNAs (Table II). The decrease was small (about 30%) but reproducible, and this suggests that the 5' end of 16S rRNA could interact with mRNA. The deletion at the 5' end of 16S rRNA reduced the miscoding by the ribosomes (Table III), restricting the MS2
RNA
A
B
T 7 Io1~ m R N A $
C
D
A kDa -66 -45
-31
B
C
D kDa -66 - 45 - 31 °21.5
-21.5
-14.4
- 14.4
Fig. 3. Fluorograms of [35S]methionine-labeled cell-free translational products obtained under the direction of MS2 RNA or T7 late mRNAs with 30S subunits reconstituted with natural 16S rRNA (A), full-length synthetic 16S rRNA (B), synthetic 16S rRNA lacking five bases at the 5' end (C), and with native 30S subunits (D).
101 TABLE II
Binding of 1"7 late mRNAs to reconstituted 30S subunits The amount of [35S]mRNA retained on the filters was determined as described under Materials and Methods with 0.16 A260 unit (A) or 0.32 A260 unit (B) of 30S subunits. The input of mRNA was 1547 800 clam in 1 A2eo unit. Results are the means of four assays with independent preparations of 30S subunits. Standard deviation of the means was less than or equal to 15%. Blanks without 30S subunits have been subtracted. No retention of mRNA was observed with 50S subunits. The retention observed with native 30S subunits was only 17 449 (A) and 36870 cpm (B). Chang and Craven also observed that native 30S subunits bind less mRNA than do reconstituted 30S subunits [41]. They relate this observation to the fact that various ribosomal proteins which may promote mRNA binding are present in fractional amounts in native 30S subunits but in unit amounts in reconstituted 30S subunits. Origin of the 16S rRNA
T7 late [3SS]mRNAs retained on the filters (cpm) A
B
Natural 16S rRNA (isolated from native 30S subunits)
49 304
100 726
Synthetic fnll-length 16S rRNA
52 985
102488
Synthetic truncated 16S rRNA (-5 nucleotides)
35049
68010
stimulation of misreading caused by streptomycin. The restriction of misreading was moderate (about 30%), compared to the strong restriction observed with riboTABLE III
Stimulation of misreading by streptomycin with reconstituted 30S subunits The error frequency is the ratio of isoleucine incorporated per phenylalanine under the direction of poly(U). It was measured as described under Materials and Methods. Each sample (100/~1) contained 0.16 A2~o unit of 30S subunits and 0.33 A2~ unit of 50S subunits. Phenylalanine (5 Ci/mmol) and isoleucine (52 Ci/mmol) were at concentrations of 10 #M and 1/~M, respectively. Streptomycin was added at a molar ratio of 5 per 30S subunit. Incubation was for 60 rain at 37°C. Results are the means of four experiments with independent preparations. The incorporation of isoleucine in the absence of streptomycin was 2000+500 clam with the different types of reconstituted 30S subunit. Standard deviation of the means of the error frequency was inferior or equal to 12%. The error frequency of 30S subunits isolated from wild-type streptomycin-sensitive E. coil was 3.0.10 -2 and that of 30S subunits isolated from a spontaneous streptomycin-resistant mutant with altered protein S12 was 0.1.10 -2. Poly(U)-directed incorporation of:
Error frequency
isoleucine (clam)
phenylalanine (cpm)
( x 102)
Natural 16S rRNA (isolated from native 30S subunits)
64 675
203 470
3.1
Synthetic fuU-length 16S rRNA
36 872
123 464
2.9
Origin of the 16S rRNA
Synthetic truncated 16S rRNA (-5 nucleotides)
20601
100748
2.0
somes from a streptomycin-resistant mutant harboring an altered protein S12. Nevertheless, it clearly indicates that the 5' end of 16S rRNA is involved in the response of the ribosome to streptomycin. Discussion
We have shown that a deletion of five bases at the 5' end of 16S rRNA decreases the binding of mRNA to 30S subunits, suggesting that this region interacts with mRNA. As indicated in the Introduction, the 5 ' e n d of 16S rRNA is located at the rear of the 30S subunit, in the neighborhood of proteins $4, $5 and $12 (see Fig. 2). The pathway followed by the mRNA on the ribosome has been characterized: it extends along the interior side of the platform, in the cleft where the decoding site is located [42]. The 'outgoing region', 5' distal to the decoding site, is located in the head of the 30S subunit where the messenger can be crosslinked to protein $7 [43]. Recent studies from Lasater et al. [11] suggest that the incoming part of the messenger leads from the cleft towards the back of the 30S subunits where it interacts with the 530 loop. Our suggestion that the 5' end of 16S rRNA also interacts with the messenger further delineates the pathway followed by the messenger towards the bottom of the 30S subunit (Fig. 4). Our results suggest that the 5' end region of 16S rRNA is involved in mRNA binding but do not indicate whether the interaction is sequence-specific, as previously suggested [7,8]. The introduction of base substitutions in the 5' end region could provide further insight on this subject. Although the deletion of five bases at the 5' end region of 16S rRNA reduced the binding of mRNA, it neither affected the efficiency of translation by the 30S subunits nor the selection of the translational start sites. A similar situation was observed when suppressing the Shine-Dalgarno interaction by mutating either mRNA or 16S rRNA [33,34]. As stressed by Gualerzi and co-workers [34,44], interactions which stabilize the bind-
5)
Fig. 4. Pathway followed by mRNA on the 30S subunit, as determined by immune electron microscopy [11,42], by crosslinkin8 studies [43] and in this study (see the text).
102
ing of mRNA to the 30S subunits may contribute to increase the concentration of the initiator triplet in the decoding area, near the binding site of the initiator tRNA but, ultimately, it is the interaction between the initiator triplet and the anticodon of the initiator tRNA which constitutes the rate-limiting step and determines the efficiency and the fidelity of initiation. The effects of interactions between 16S rRNA and mRNA are not readily detectable in vitro but it is most probable that they confer a selective advantage in the competitive context that prevails in vivo. A deletion of five nucleotides at the 5' end of 16S rRNA restricted the response of the ribosome to streptomycin. This drug binds to 16S rRNA in the 915 region, located at the convergence of the three main domains of 16S rRNA [6,45], where it protects residues 911-915 and, to a lesser extent, residue 909 and the proximal residues 1413, 1487 and 1494 (Fig. 1). The recent models describing the folding of 16S rRNA inside the 30S subunit [5,18,19] locate the 915 area on the 50S subunit side of the 30S subunit, lining the decoding site, and in the vicinity of the 5' t,,rminal helix and the 17-19/916-918 pseudoknot helix. The deletion which we have introduced at the 5' end of 16S rRNA flanks the 5' terminal helix, but on the opposite side of the subunit from the decoding center (see Fig. 2). It is now accepted that the control of translational accuracy and of the response to streptomycin is exerted by 16S rRNA through conformational changes which are modulated by proteins $4, $5 and $12 (Refs. 46,47 and Refs. therein). It is probable that the 5' end of 16S rRNA is involved in the modulation exerted by these proteins and that conformational changes are transmitted from that region to the decoding center through the 5' terminal and the pseudoknot helices. A deletion in the 5' end region would perturb the transmission of these conformational changes. Interestingly, Harris et al. [48] have recently observed that a single base change in the chloroplast 16S rRNA of Chlamydomonas reinhardtii, at a position equivalent to position 13 in E. coli 16S rRNA, confers resistance to streptomycin. This observation further supports the involvement of the 5' end region of 16S rRNA in the response of the ribosome to streptomycin. Stern et al. [5] proposed that streptomycin induces a tightening in the 915 region, which decreases the probability of rejection of non-cognate tRNA. It is tempting to suggest that the deletion at the 5' end of 16S rRNA interferes with this tightening. Acknowledgements We thank Drs. Guy Boileau, Gabriel Gingras and Micha61 Laughrea for valuable discussions and comments. This work was supported by the Medical Research Council of Canada. Daniel Leclerc is the recipient of a studentship from the Medical Research Council
of Canada. We are grateful to Ms. Danielle Vinet for typing this manuscript and Mr. Yves Villeneuve for preparing the figures. References 1 De Smit, M.H. and Van Duin, J. (1990) Prog. Nucleic Acid Res. Mol. Biol. 38, 1-35. 2 Van Knippenberg, P.H. (1990) in The Structure, Function and Evolution of Ribosomes (Hill, W.E., ed.), American Society for Microbiology, Washington, D.C., in press. 3 Hershey, J.W.B. (1987) in Escherichia coli and Salmonella typhimurium (Neidhardt, F.C., Ingraham, J.L., Brooks, K.L., Magasanik, B., Schaechter, M. and Umbarger, H.E., eds.), pp. 613-647, American Society for Microbiology, Washington, D.C. 4 Gold, L. (1988) Annu. Rev. Biochem. 57, 199-233. 5 Stern, S., Weiser, B. and Noller, H.F. (1988) J. Mol. Biol. 204, 447-481. 6 Moazed, D. and Noller, H.F. (1987) Nature 327, 389-394. 7 Petersen, G.B., Stockwell, P.A. and Hill, D.F. (1988) EMBO J. 7, 3957-3962. 8 Van Knippenberg, P.H. (1975) Nucleic Acids Res. 2, 79-85. 90akes, M., Henderson, E., Scheinman, A., Clark, M. and Lake, J.A. (1986) In Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 47-67, Springer-Verlag, New York. 10 Trempe, M.R., Ohgi, K. and Glitz, D.G. (1982) J. Biol. Chem. 257, 9822-9829. 11 Lasater, L.S., Montesano-Roditis, L., Cann, P.A. and Glitz, D.G. (1990) Nucleic Acids Res. 18, 477-485. 12 Oakes, M.I. and Lake, J.A. (1990) J. Mol. Biol. 211, 897-906. 13 Mochalova, L.V., Shatsky, I.N., Bogdanov, A.A. and Vasiliev, V.D. (1982) J. Mol. Biol. 159, 637-650. 14 Gornicki, P., Nurse, K , Hellmann, W., Boubiik, M. ~ c i Ofengand, J. (1984) J. Biol. Chem. 259, 10493-10498. 15 Oakes, M.I., Clark, M.W., Henderson, E. and Lake, J.A. (1986) Proc. Natl. Acad. Sci. USA 83, 275-279. 16 Capel, M.S., Engelman, D.M., Freeborn, B.R., Kjeldgaard, M., Langer, J.A., Ramakrishnan, V., Schindler, D.G., Schneider, D.K., Schoenborn, B.P., Sillers, I.Y., Yabuki, S. and Moore, P.B. (1987) Science 238, 1403-1406. 17 St~Sffler, G. and St/Sffler-Meilicke, M. (1986) in Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 28-46, Springer-Verlag, New York. 18 Brimacombe, R., Atmadja, J., Stiege, W. and Schiiler, D. (1988) J. Mol. Biol. 199, 115-136. 19 Oakes, M.I., Kahan, L. and Lake, J.A. (1990) J. Mol. Biol. 211, 907-918. 20 Towbin, H. and Elson, D. (1978) Nucleic Acids Res. 5, 3389-3407. 21 Broude, N.E., Kussova, K.S., Medvedeva, N.I. and Budowsky, E.I. (1983) Eur. J. Biochem. 132, 139-145. 22 Gimautdinova, O.I., Zenkova, M.A., Karpova, G.G. and Podust, L.M. (1984) Mol. Biol. (Moscow) 18, 734-744. 23 Cooperman, B.S. (1987) Pharmacol. Ther. 34, 271-302. 24 Stern, S., Powers, T., Changchien, L.M. and Noller, H.F. (1988) J. Mol. Biol. 201,683-695. 25 Osswald, M., Greuer, B., Brimacombe, R., St/Sffler, G., Baiimert, H. and Fasold, H. (1987) Nucleic Acids Res. 15, 3221-3240. 26 Vazquez, D. (1979) Mol. Biol. Biochem. Biophys. 30, 89-95. 27 Wallace, B.J., Tai, P.C. and Davis, B.D. (1979) in Antibiotics V Mechanism of Action of Antibacterial Agents (Hahn, F.E., ed.), pp. 272-303, Springer-Verlag, Berlin. 28 Brakier-Gingras, L. and Phoenix, P. (1984) Can. J. Biochem. Cell. Biol. 62, 231-244. 29 Krzyzosiak, W., Denman, R., Nurse, K., Hellmann, W., Boublik, M., Gehrke, C.W., Agris, P.F. and Ofengand, J. (1987) Biochemistry 26, 2353-23641 ......
103 30 Melanin, P., Gravel, M., Boileau, G. and Brakier-Gingras, L. (1987) Biochem. Cell Biol. 65, 1022-1030. 31 Gravel, M., Leclerc, D., Melanc,xon, P. and Brakier-Gingras, L. (1989) Nucleic Acids Res. 17, 2723-2732. 32 Leclerc, D. and Brakier-Gingras, L. (1990) Biochem. Cell Biol. 68, 169-179. 33 Melanc,son, P., Leclerc, D., Destroismaisons, N. and BrakierGingras, L. (1990) Biochemistry 29, 3402-3407. 34 Calogero, R.A., Pon, C.L., Canonaco, M.A. and Gualerzi, C.O. (1988) Proc. Natl. Acad. Sci. USA 85, 6427-6431. 35 Taylor, J.W., Schmidt, W., Cosstick, R., Okruszek, A. and Eckstein, F. (1985) Nucleic Acids Res. 13, 8749-8764. 36 Taylor, J.W., Ott, J. and Eckstein, F. (1985) Nucleic Acids Res. 13, 8765-8785. 37 Davanloo, P., Rosenberg, A.H., Durra, J.J. and Studier, F.W. (1984) Proc. Natl. Acad. Sci. USA 81, 2035-2039. 38 Melton, D.A., Krieg, P.A., Rebagliati, M.R., Maniatis, T., Zinn, K. and Green, M.R. (1984) Nucleic Acids Res. 12, 7035-7056. 39 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.
40 Gellebter, J. (1987) Focus 9, 5-8. 41 Chang, C. and Craven, G.R. (1977) J. Mol. Biol. 117, 401-418. 42 Olson, H.M., Lasater, L.S., Cann, P.A. and Glitz, D.G. (1988) J. Biol. Chem. 263, 15196-15204. 43 Stade, K., Rinke-Appel, J. and Brimacombe, R. (1989) Nucleic Acids Res. 17, 9889-9908. 44 Gualerzi, C.O., Pon, C.L., Pawlik, R.T., Canonaco, M.A., Paci, M. and Wintermeyer, W. (1986) in Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 621-641, Springer-Verlag, New York. 45 Gravel, M., Melan~on, P. and Brakier-Gingras, L. (1987) Biochemistry 26, 6227-6232. 46 Allen, P.N. and Noller, H.F. (1989) J. Mol. Biol. 208, 457-468. 47 Stem, S., Powers, T., Changchien, L.M. and Noller, H.F. (1989) Science 244, 783-790. 48 Harris, E.H., Burkhart, B.D., Gilham, N.W. and Boynton, J.E. (1989) Genetics 123, 281-292.
Biochimica et Biophysica Acta, 1050 (1990) 98-103 Elsevier
BBAEXP 92098
A deletion mutation at the 5' end of Escherichia coli 16S ribosomal RNA Pierre Melan~on, Daniel Leclerc and Lra Brakier-Gingras D~partement de Biochimie, Universit~de Montreal, Montreal, Quebec (Canada) (Received 16 May 1990)
Key words: ribosomal RNA; 16S; Translation initiation; Translation accuracy
A deletion of five nucleotides was introduced at the 5' end of the Escherichia coli 16S rRNA gene cloned in an appropriate vector under control of a T7 promoter. The 16S rRNA generated by in vitro transcription could be assembled into 30S subunits. The deletion did not affect the efficiency of translation of natural messengers and the correct selection of the reading frame. However, it reduced the binding of the messengers, which suggests that the 5' end of 16S rRNA is located on the pathway followed by the messengers on the 30S subunits. The deletion also restricted the stimulation of misreading by streptomycin in a poly(U)-directed system. This is in accord with the proximity of the 5' end of 168 rRNA to proteins 84, $5 and S12, which are known to be involved in the control of translational accuracy.
Introduction The process of translation initiation involves the selection by the 30S ribosomal subunits of an initiator codon on the messenger RNA. This codon is AUG in about 90% of Escherichia coli genes. A critical determinant for the efficiency of the initiation step is the lack of secondary structure in the mRNA around the initiator codon. Various interactions have been proposed to stabilize the binding of the messenger to the 30S subunits. The best-characterized is the interaction between the Shine-Dalgarno sequence, a purine-rich sequence on mRNA, upstream of the initiator AUG codon, and the anti-Shine-Dalgarno region, a complementary pyrimidine-rich sequence near the 3' end of 16S rRNA (reviewed in Refs. 1-4). The 5' end region of 16S rRNA (Fig. 1) is a likely candidate for an additional interaction with sequences downstream from the initiator codon, at the beginning of the coding region of mRNA [7,8]. The 5' end of 16S rRNA has been located at the rear of the 30S subunit, below the 530 loop (Fig. 2). It is distal from the decoding region in the cleft, and is proximal to proteins $4, $5 and S12. Affinity labeling studies (Refs. 20-22 and reviewed in Ref. 23) have shown that proteins $4, $5 and S12 can be
Correspondence: L. Brakier-Gingras, Drpartement de Biochimie, Universit6 de Montrral, Montrral H3C 3J7, Canada.
linked to the messengers. Moreover, crosslinking and footprinting studies demonstrate a direct interaction between the 5' end of 16S rRNA and proteins $4 and $5 [24,25]. These observations support an interaction between the 5' end of 16S rRNA and mRNA. Proteins $4, $5 and S12 are also known to be involved in the control of translational accuracy and in the response to streptomycin, an error-promoting aminoglycoside antibiotic (reviewed in Refs. 26-28). Their proximity to the 5' end region of 16S rRNA also suggests that this region could play a role in the control of translational fidelity, in addition to interacting with mRNA. We and others have recently developed a method which functions entirely in vitro to investigate the consequences of site-directed mutagenesis in E. coli 16S rRNA on ribosome function (Refs. 29-31 and reviewed in Ref. 32). In this system, mutations are introduced into the 16S rRNA gene cloned in an appropriate vector, directly downstream from a "1"7 promoter and flanked at its 3' end by an appropriate restriction site. In vitro transcribed mutated 16S rRNA is then assembled into 30S subunits, which are assayed for their protein synthesis activity. Using this system, we have previously shown that suppressing the Shine-Dalgarno interaction by deleting the anti-Shine-Dalgarno region at the 3' end of 16S rRNA neither affected the capacity of the 30S subunits to translate mRNA nor their aptitude to select the translation start sites correctly [33]. This was in full agreement with a study by Calogero et
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
99 530 LOOP
Materials and Methods
,°CC°cG 530
C (3
520-ACc
G/
G-C G-C CU GU
A AU GU__A. U
5'ko c - o . \ U--A
_ \
~ k 31~ .
u~Ac-c-54o cGG,,-~
~A G
~-~ A -- U 550
u',% A c-G
B901
uAC GGGA~uACGGccGc 2
o ~ ~ ""U / G 920
9 1 5 AREA G =~,v
c uCGG.AC'AA~C^CaO~Q~---~ .~^C.,,.,U..4" """'"
,~G A A
\uU~u u
~;AGu UGa \
^1o
C
ACUGUUCCG~ A
A
UU" AG--C G ~ A^',, A .... l A ^C ~,,A t, 30 , G4-.-~ ~UAt; 500 . t , r_t, U/
C C~1¢00 ,.,
'~'~'AUGUC C U G~ 400~C~ \_G G CU, GC~40
A A5 A I AJ
t, C C G
5'
u
/
:c
A
-®u
A A
C G
C--G U 14.10---A -- U-1490 A C --G A-15OO CmG C
®'@
U.G G.u G--C
A
A G
Fig. 1. Secondary structure of a portion of E. coli 165 rRNA [51, showing the 5' terminal region and parts of the different domains. The extent of the deletion introduced at the 5' end is indicated by the arrows. The circled nucleotides are protected against chemical probes by streptomycin [61.
al. [34], who did not find any difference when comparing the translation of two mRNA, which were identical except for the presence or absence of the Shine-Dalgarno sequence. In the present study, we have investigated how deleting five bases at the 5' end of 16S rRNA affects the activity of ribosomes.
Cleft
(decoding center)\c--7
/
Fig. 2. Model of the 305 sub~_it as seen from the solvent side [91. The positions of the 530 loop [10-121, the 5' end of the 16S rRNA [131, and the decoding center [14,15] have been mapped by immune and DNA-hybddizafion electron microscopy. Proteins S4, $5, $7 and S12 have been mapped by neutron diffraction [16] and immune electron microscopy [9,17]. The dotted area indicates the position of the 5' terminal and the pseudoknot helices of 16S rRNA at the interface with the 50S subunit [5,18,19].
Construction of the transcription vectors and preparation of the 16S rRNA transcripts The transcription vector pDI.A417 is a derivative of the phagemid Bluescript SK- (Stratagene). A first derivative of Bluescript (BLM1) was constructed by creating a NdeI site, directly upstream from the T7 promoter. The phagemid pDL4417 was then obtained by inserting between this Nde I site and the Xba I site in the polylinker, an NdeI-XbaI fragment from p P M l l 4 [31], which contains the E. coli 16S rRNA gene preceded by a T7 promoter. Two GC pairs have been inserted between the end of the T7 promoter and the beginning of the 16S rRNA gene. The 3' end of the 16S rRNA gene is part of a Bsu36I site, immediately followed by a XbaI site. A deletion mutant of pDL4417, pPM5'-A5, lacking 5 nucleotides at the 5' end of the 16S rRNA gene, was obtained by the phosphorothioate method of Eckstein [35,36], using the Amersham in vitro oligonucleotide-directed mutagenesis system. In vitro transcription of pDL4417 or its deletion derivative, linearized with Bsu36I (New England Biolabs), was carried out following standard procedures [37,38]. As a consequence of the construction of the plasmids, all 16S rRNA transcripts contain two additional GC residues at the 5' end. The 3' end is identical to that of natural 16S rRNA [31]. Assembly of 30S subunits The 30S subunits were assembled with either the full-length 16S rRNA transcript or its deletion derivative, as described previously [31,33]. The 5' end of 16S rRNA was analyzed by primer extension dideoxy sequencing [39] after phenol-extraction of the RNA from the reassembled subunits. Primer extension was done with reverse transcriptase (Pharmacia), using a (5'32p)-labeled synthetic deoxynucleotide primer [40], complementary to bases 31-45 in the natural 16S rRNA. Primer extension was also carded out without the dideoxy nucleotides in order to assess the amount of contaminating natural 16S rRNA in the reconstituted 30S subunits. Protein synthesis assays Polypeptide synthesis was programmed either with MS2 RNA or with the late mRNAs from the T7 phage. The reaction was carried out as described [31,33], using 0.16 A260 unit of 30S subunits and 0.33 A260 unit of 50S subunits. The translation products were analyzed by one-dimensional polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate as described [33]. Misreading assays under the direction of poly(U) were as described [31]: the cognate amino acid, [12C]phenylalanine, and the non-cognate amino acid, [3H]isoleucine (52 Ci/mmol), were added simulta-
100 neously at of concentrations of 10 mM and 1 mM, respectively. Streptomycin was added at a molar ratio of 5 per ribosome. In parallel assays, the incorporation of the cognate amino acid, [3H]phenylalanine (5 Ci/mmol), was measured in the presence of [12Clisoleucine and streptomycin. Binding of messenger RNA to 30S subunits The late mRNAs from the T7 phage were labeled with 35S during in vitro transcription, by including 25 #Ci of [a-3~S]UTP (Amersham; 1250 Ci/mmol) to 1 ml of the transcription mixture. Binding assays were performed by Millipore filtration, using 0.16 or 0.32 A260 unit of 30S subunits and 1 A260 unit of T7 messengers. The complex between the subunits and the mRNA, (200 #1), was incubated for 15 min at 37°C in a 50 mM Tris-HCl buffer (pH 7.8) containing 7.5 mM magnesium acetate, 100 mM NH4C1 and 1 mM dithiothreitol, then diluted to 3 ml with the same buffer, filtered through Millipore nitrocellulose filters (HA-0.22 m#) and washed three times with 3 ml of a cold buffer containing 10 mM Tris-HCl (pH 7.5), 10 mM magnesium acetate, 60 mM NH4C1 and 5 mM fl-mercaptoethanol. Filter-bound 355 w a s measured by liquid scintillation counting. Values were corrected for background retention of mRNA in the absence of 30S subunits. The specificity of the binding of mRNA was assessed by verifying the absence of retention with 50S subunits. Results
The 30S subunits were assembled with the full-length 16S rRNA transcript or its deletion derivative lacking five nucleotides at the 5' end. The reconstituted 30S particles comigrated with native 30S subunits in a sucrose gradient and analysis by two-dimensional gel electrophoresis indicated that they contained the full complement of ribosomal proteins. RNA sequencing analysis confirmed that the 5' end of the 16S rRNA transcripts had the expected sequence and primer extension assays indicated that the transcripts represented 95% or more of the 16S rRNA in the reconstituted subunits, the remaining 16S rRNA being natural 16S rRNA brought in as a contaminant of the ribosomal proteins (data not shown). The deletion at the 5' end of the 16S rRNA did not significantly affect the capacity of the 30S subunits to translate natural mRNAs, such as MS2 RNA or T7 phage late mRNAs (Table I). Furthermore, kinetic studies revealed no difference between mutated and unmutated 30S subunits during the time course of the reaction (not shown). The proteins synthesized by the mutated and m u t a t e d ribosomes were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis. The patterns of the trallslation products were identical whether the 30S subunits contained the full-length or the mutant 16S rRNAs (Fig. 3): a single
TABLE I Protein synthesis activity of reconstituted 30S subunits The activity of reconstituted 30S subunits was assessed by measuring the incorporation of [3H]leucine (1.9 Ci/mmol) and [3Hlvaline (1.8 Ci/mmol) into trichloroacetic acid-insoluble material under the direction of MS2 RNA or T7 late mRNAs. Each sample (100 ~1) contained 0.16 A260 unit of 30S subunits and 0.33 A260 unit of 50S subunits. Incubation was for 60 min at 37°C. Results are the means of four assays with independent preparations of 30S subunits. Standard deviation of the means was less than or equal to 17%. Blanks without 30S subunits have been subtracted. Origin of the 16S rRNA
Activity (in clam) with MS2 RlqA
with T7 late mRNAs
Natural 16S rRNA (isolated from native 30S subunits)
8 648
12 490
Synthetic full-length 16S rRNA
7160
10208
Synthetic truncated 16S rRNA (-5 nucleotides)
6 668
9 346
band corresponding to the coat protein was seen in all cases with MS2 RNA. A more complex pattern was observed with T7 phage late mRNAs, but the sizes and relative amounts of the synthesized proteins were identical with all types of 30S subunit. These results indicate that the deletion did not prevent the correct selection of the reading frame of translation and that the 5' end region of 16S rRNA does not appear essential for protein synthesis. However, the binding of mRNA to the mutated 30S subunits lacking five bases at the 5' end of 16S rRNA was decreased, as shown by the filtration assays measuring the binding of labeled T7 late mRNAs (Table II). The decrease was small (about 30%) but reproducible, and this suggests that the 5' end of 16S rRNA could interact with mRNA. The deletion at the 5' end of 16S rRNA reduced the miscoding by the ribosomes (Table III), restricting the MS2
RNA
A
B
T 7 Io1~ m R N A $
C
D
A kDa -66 -45
-31
B
C
D kDa -66 - 45 - 31 °21.5
-21.5
-14.4
- 14.4
Fig. 3. Fluorograms of [35S]methionine-labeled cell-free translational products obtained under the direction of MS2 RNA or T7 late mRNAs with 30S subunits reconstituted with natural 16S rRNA (A), full-length synthetic 16S rRNA (B), synthetic 16S rRNA lacking five bases at the 5' end (C), and with native 30S subunits (D).
101 TABLE II
Binding of 1"7 late mRNAs to reconstituted 30S subunits The amount of [35S]mRNA retained on the filters was determined as described under Materials and Methods with 0.16 A260 unit (A) or 0.32 A260 unit (B) of 30S subunits. The input of mRNA was 1547 800 clam in 1 A2eo unit. Results are the means of four assays with independent preparations of 30S subunits. Standard deviation of the means was less than or equal to 15%. Blanks without 30S subunits have been subtracted. No retention of mRNA was observed with 50S subunits. The retention observed with native 30S subunits was only 17 449 (A) and 36870 cpm (B). Chang and Craven also observed that native 30S subunits bind less mRNA than do reconstituted 30S subunits [41]. They relate this observation to the fact that various ribosomal proteins which may promote mRNA binding are present in fractional amounts in native 30S subunits but in unit amounts in reconstituted 30S subunits. Origin of the 16S rRNA
T7 late [3SS]mRNAs retained on the filters (cpm) A
B
Natural 16S rRNA (isolated from native 30S subunits)
49 304
100 726
Synthetic fnll-length 16S rRNA
52 985
102488
Synthetic truncated 16S rRNA (-5 nucleotides)
35049
68010
stimulation of misreading caused by streptomycin. The restriction of misreading was moderate (about 30%), compared to the strong restriction observed with riboTABLE III
Stimulation of misreading by streptomycin with reconstituted 30S subunits The error frequency is the ratio of isoleucine incorporated per phenylalanine under the direction of poly(U). It was measured as described under Materials and Methods. Each sample (100/~1) contained 0.16 A2~o unit of 30S subunits and 0.33 A2~ unit of 50S subunits. Phenylalanine (5 Ci/mmol) and isoleucine (52 Ci/mmol) were at concentrations of 10 #M and 1/~M, respectively. Streptomycin was added at a molar ratio of 5 per 30S subunit. Incubation was for 60 rain at 37°C. Results are the means of four experiments with independent preparations. The incorporation of isoleucine in the absence of streptomycin was 2000+500 clam with the different types of reconstituted 30S subunit. Standard deviation of the means of the error frequency was inferior or equal to 12%. The error frequency of 30S subunits isolated from wild-type streptomycin-sensitive E. coil was 3.0.10 -2 and that of 30S subunits isolated from a spontaneous streptomycin-resistant mutant with altered protein S12 was 0.1.10 -2. Poly(U)-directed incorporation of:
Error frequency
isoleucine (clam)
phenylalanine (cpm)
( x 102)
Natural 16S rRNA (isolated from native 30S subunits)
64 675
203 470
3.1
Synthetic fuU-length 16S rRNA
36 872
123 464
2.9
Origin of the 16S rRNA
Synthetic truncated 16S rRNA (-5 nucleotides)
20601
100748
2.0
somes from a streptomycin-resistant mutant harboring an altered protein S12. Nevertheless, it clearly indicates that the 5' end of 16S rRNA is involved in the response of the ribosome to streptomycin. Discussion
We have shown that a deletion of five bases at the 5' end of 16S rRNA decreases the binding of mRNA to 30S subunits, suggesting that this region interacts with mRNA. As indicated in the Introduction, the 5 ' e n d of 16S rRNA is located at the rear of the 30S subunit, in the neighborhood of proteins $4, $5 and $12 (see Fig. 2). The pathway followed by the mRNA on the ribosome has been characterized: it extends along the interior side of the platform, in the cleft where the decoding site is located [42]. The 'outgoing region', 5' distal to the decoding site, is located in the head of the 30S subunit where the messenger can be crosslinked to protein $7 [43]. Recent studies from Lasater et al. [11] suggest that the incoming part of the messenger leads from the cleft towards the back of the 30S subunits where it interacts with the 530 loop. Our suggestion that the 5' end of 16S rRNA also interacts with the messenger further delineates the pathway followed by the messenger towards the bottom of the 30S subunit (Fig. 4). Our results suggest that the 5' end region of 16S rRNA is involved in mRNA binding but do not indicate whether the interaction is sequence-specific, as previously suggested [7,8]. The introduction of base substitutions in the 5' end region could provide further insight on this subject. Although the deletion of five bases at the 5' end region of 16S rRNA reduced the binding of mRNA, it neither affected the efficiency of translation by the 30S subunits nor the selection of the translational start sites. A similar situation was observed when suppressing the Shine-Dalgarno interaction by mutating either mRNA or 16S rRNA [33,34]. As stressed by Gualerzi and co-workers [34,44], interactions which stabilize the bind-
5)
Fig. 4. Pathway followed by mRNA on the 30S subunit, as determined by immune electron microscopy [11,42], by crosslinkin8 studies [43] and in this study (see the text).
102
ing of mRNA to the 30S subunits may contribute to increase the concentration of the initiator triplet in the decoding area, near the binding site of the initiator tRNA but, ultimately, it is the interaction between the initiator triplet and the anticodon of the initiator tRNA which constitutes the rate-limiting step and determines the efficiency and the fidelity of initiation. The effects of interactions between 16S rRNA and mRNA are not readily detectable in vitro but it is most probable that they confer a selective advantage in the competitive context that prevails in vivo. A deletion of five nucleotides at the 5' end of 16S rRNA restricted the response of the ribosome to streptomycin. This drug binds to 16S rRNA in the 915 region, located at the convergence of the three main domains of 16S rRNA [6,45], where it protects residues 911-915 and, to a lesser extent, residue 909 and the proximal residues 1413, 1487 and 1494 (Fig. 1). The recent models describing the folding of 16S rRNA inside the 30S subunit [5,18,19] locate the 915 area on the 50S subunit side of the 30S subunit, lining the decoding site, and in the vicinity of the 5' t,,rminal helix and the 17-19/916-918 pseudoknot helix. The deletion which we have introduced at the 5' end of 16S rRNA flanks the 5' terminal helix, but on the opposite side of the subunit from the decoding center (see Fig. 2). It is now accepted that the control of translational accuracy and of the response to streptomycin is exerted by 16S rRNA through conformational changes which are modulated by proteins $4, $5 and $12 (Refs. 46,47 and Refs. therein). It is probable that the 5' end of 16S rRNA is involved in the modulation exerted by these proteins and that conformational changes are transmitted from that region to the decoding center through the 5' terminal and the pseudoknot helices. A deletion in the 5' end region would perturb the transmission of these conformational changes. Interestingly, Harris et al. [48] have recently observed that a single base change in the chloroplast 16S rRNA of Chlamydomonas reinhardtii, at a position equivalent to position 13 in E. coli 16S rRNA, confers resistance to streptomycin. This observation further supports the involvement of the 5' end region of 16S rRNA in the response of the ribosome to streptomycin. Stern et al. [5] proposed that streptomycin induces a tightening in the 915 region, which decreases the probability of rejection of non-cognate tRNA. It is tempting to suggest that the deletion at the 5' end of 16S rRNA interferes with this tightening. Acknowledgements We thank Drs. Guy Boileau, Gabriel Gingras and Micha61 Laughrea for valuable discussions and comments. This work was supported by the Medical Research Council of Canada. Daniel Leclerc is the recipient of a studentship from the Medical Research Council
of Canada. We are grateful to Ms. Danielle Vinet for typing this manuscript and Mr. Yves Villeneuve for preparing the figures. References 1 De Smit, M.H. and Van Duin, J. (1990) Prog. Nucleic Acid Res. Mol. Biol. 38, 1-35. 2 Van Knippenberg, P.H. (1990) in The Structure, Function and Evolution of Ribosomes (Hill, W.E., ed.), American Society for Microbiology, Washington, D.C., in press. 3 Hershey, J.W.B. (1987) in Escherichia coli and Salmonella typhimurium (Neidhardt, F.C., Ingraham, J.L., Brooks, K.L., Magasanik, B., Schaechter, M. and Umbarger, H.E., eds.), pp. 613-647, American Society for Microbiology, Washington, D.C. 4 Gold, L. (1988) Annu. Rev. Biochem. 57, 199-233. 5 Stern, S., Weiser, B. and Noller, H.F. (1988) J. Mol. Biol. 204, 447-481. 6 Moazed, D. and Noller, H.F. (1987) Nature 327, 389-394. 7 Petersen, G.B., Stockwell, P.A. and Hill, D.F. (1988) EMBO J. 7, 3957-3962. 8 Van Knippenberg, P.H. (1975) Nucleic Acids Res. 2, 79-85. 90akes, M., Henderson, E., Scheinman, A., Clark, M. and Lake, J.A. (1986) In Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 47-67, Springer-Verlag, New York. 10 Trempe, M.R., Ohgi, K. and Glitz, D.G. (1982) J. Biol. Chem. 257, 9822-9829. 11 Lasater, L.S., Montesano-Roditis, L., Cann, P.A. and Glitz, D.G. (1990) Nucleic Acids Res. 18, 477-485. 12 Oakes, M.I. and Lake, J.A. (1990) J. Mol. Biol. 211, 897-906. 13 Mochalova, L.V., Shatsky, I.N., Bogdanov, A.A. and Vasiliev, V.D. (1982) J. Mol. Biol. 159, 637-650. 14 Gornicki, P., Nurse, K , Hellmann, W., Boubiik, M. ~ c i Ofengand, J. (1984) J. Biol. Chem. 259, 10493-10498. 15 Oakes, M.I., Clark, M.W., Henderson, E. and Lake, J.A. (1986) Proc. Natl. Acad. Sci. USA 83, 275-279. 16 Capel, M.S., Engelman, D.M., Freeborn, B.R., Kjeldgaard, M., Langer, J.A., Ramakrishnan, V., Schindler, D.G., Schneider, D.K., Schoenborn, B.P., Sillers, I.Y., Yabuki, S. and Moore, P.B. (1987) Science 238, 1403-1406. 17 St~Sffler, G. and St/Sffler-Meilicke, M. (1986) in Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 28-46, Springer-Verlag, New York. 18 Brimacombe, R., Atmadja, J., Stiege, W. and Schiiler, D. (1988) J. Mol. Biol. 199, 115-136. 19 Oakes, M.I., Kahan, L. and Lake, J.A. (1990) J. Mol. Biol. 211, 907-918. 20 Towbin, H. and Elson, D. (1978) Nucleic Acids Res. 5, 3389-3407. 21 Broude, N.E., Kussova, K.S., Medvedeva, N.I. and Budowsky, E.I. (1983) Eur. J. Biochem. 132, 139-145. 22 Gimautdinova, O.I., Zenkova, M.A., Karpova, G.G. and Podust, L.M. (1984) Mol. Biol. (Moscow) 18, 734-744. 23 Cooperman, B.S. (1987) Pharmacol. Ther. 34, 271-302. 24 Stern, S., Powers, T., Changchien, L.M. and Noller, H.F. (1988) J. Mol. Biol. 201,683-695. 25 Osswald, M., Greuer, B., Brimacombe, R., St/Sffler, G., Baiimert, H. and Fasold, H. (1987) Nucleic Acids Res. 15, 3221-3240. 26 Vazquez, D. (1979) Mol. Biol. Biochem. Biophys. 30, 89-95. 27 Wallace, B.J., Tai, P.C. and Davis, B.D. (1979) in Antibiotics V Mechanism of Action of Antibacterial Agents (Hahn, F.E., ed.), pp. 272-303, Springer-Verlag, Berlin. 28 Brakier-Gingras, L. and Phoenix, P. (1984) Can. J. Biochem. Cell. Biol. 62, 231-244. 29 Krzyzosiak, W., Denman, R., Nurse, K., Hellmann, W., Boublik, M., Gehrke, C.W., Agris, P.F. and Ofengand, J. (1987) Biochemistry 26, 2353-23641 ......
103 30 Melanin, P., Gravel, M., Boileau, G. and Brakier-Gingras, L. (1987) Biochem. Cell Biol. 65, 1022-1030. 31 Gravel, M., Leclerc, D., Melanc,xon, P. and Brakier-Gingras, L. (1989) Nucleic Acids Res. 17, 2723-2732. 32 Leclerc, D. and Brakier-Gingras, L. (1990) Biochem. Cell Biol. 68, 169-179. 33 Melanc,son, P., Leclerc, D., Destroismaisons, N. and BrakierGingras, L. (1990) Biochemistry 29, 3402-3407. 34 Calogero, R.A., Pon, C.L., Canonaco, M.A. and Gualerzi, C.O. (1988) Proc. Natl. Acad. Sci. USA 85, 6427-6431. 35 Taylor, J.W., Schmidt, W., Cosstick, R., Okruszek, A. and Eckstein, F. (1985) Nucleic Acids Res. 13, 8749-8764. 36 Taylor, J.W., Ott, J. and Eckstein, F. (1985) Nucleic Acids Res. 13, 8765-8785. 37 Davanloo, P., Rosenberg, A.H., Durra, J.J. and Studier, F.W. (1984) Proc. Natl. Acad. Sci. USA 81, 2035-2039. 38 Melton, D.A., Krieg, P.A., Rebagliati, M.R., Maniatis, T., Zinn, K. and Green, M.R. (1984) Nucleic Acids Res. 12, 7035-7056. 39 Sanger, F., Nicklen, S. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463-5467.
40 Gellebter, J. (1987) Focus 9, 5-8. 41 Chang, C. and Craven, G.R. (1977) J. Mol. Biol. 117, 401-418. 42 Olson, H.M., Lasater, L.S., Cann, P.A. and Glitz, D.G. (1988) J. Biol. Chem. 263, 15196-15204. 43 Stade, K., Rinke-Appel, J. and Brimacombe, R. (1989) Nucleic Acids Res. 17, 9889-9908. 44 Gualerzi, C.O., Pon, C.L., Pawlik, R.T., Canonaco, M.A., Paci, M. and Wintermeyer, W. (1986) in Structure, Function and Genetics of Ribosomes (Hardesty, B. and Kramer, G., eds.), pp. 621-641, Springer-Verlag, New York. 45 Gravel, M., Melan~on, P. and Brakier-Gingras, L. (1987) Biochemistry 26, 6227-6232. 46 Allen, P.N. and Noller, H.F. (1989) J. Mol. Biol. 208, 457-468. 47 Stem, S., Powers, T., Changchien, L.M. and Noller, H.F. (1989) Science 244, 783-790. 48 Harris, E.H., Burkhart, B.D., Gilham, N.W. and Boynton, J.E. (1989) Genetics 123, 281-292.